This project will investigate the hypothesis that: (a) children with autism have unusual difficulty grasping the relation between one thing &another (e.g., between a stimulus object &reward object) that cannot be attributed to cognitive delay, (b) critical advances in this fundamental ability to perceive conceptual connections between physically unconnected things occur during the 2nd year of life, and (c) children with autism can show rule-learning and can grasp abstract concepts when fairly simple modifications are made in how the relation between objects is presented to them. Typically-developing (TD) infants under 21 months and children with autism have inordinate difficulty acquiring the "delayed nonmatching to sample [DNMS]" rule even with minimal delays, when the rewards are placed in shallow wells just beneath the stimuli;though once they've learned the rule, they succeed at long delays. Since children with autism appear to fail DNMS in the same ways as do infants, they may be failing for the same reasons and may succeed under the same conditions. If the reward is velcroed to the base of the stimulus (reward still hidden when stimulus is atop a well), infants of 12 months readily grasp the nonmatching rule. We predict that children with autism will similarly succeed when the rewards are attached to (though detachable from) the base of the stimuli. If children with autism are able to grasp the abstract rule and the connection between stimulus and reward in the velcro condition, the implications for intervention are exciting. Most behavioral training with children with autism has not considered whether cue and referent are physically connected. The ability to form abstract concepts and grasp the relation between cue and referent is critical for diverse facets of cognitive and social development. Thus, we expect our findings will have direct and immediate implications for teaching some concepts to some young children with autism more efficiently, earlier, and with greater real acquisition than previously thought possible. To better understand the ability of children with autism to relate one thing to another, and the importance of physical connectivity to their ability to do that, 3 tasks, each with 2-3 conditions, will be used. To determine whether this is specific to children with autism, at what developmental level children with autism are performing, and the typical developmental progression in the unfolding of this ability, these tasks will be given to mildly developmentally-delayed preschoolers with (a) autism and (b) Down syndrome (each group with a mean age of 45 months and mean mental age of 27 months), and TD children of 27, 24, 21,18, 15, and 12 months.